Files
sft-qwen2.5-1.5b/eval-results/mbpp/metrics.json

11 lines
210 B
JSON
Raw Normal View History

{
"mbpp": {
"pass@1": {
"num_entries": 378,
"avg_tokens": 66,
"gen_seconds": 20,
"passing_base_tests": 60.58201058201058,
"passing_plus_tests": 51.851851851851855
}
}
}